AND,,,,,

Raj Nambiar has said:

"What I meant was that, if the output
         impedence of the source that drives 
         7135 input is 1Meg

Raj: Please also try to change this problem. Trying to use a 1 Mohm source impedance on a circuit board into an A/D converter is almost fool hardy. That is where your problems are. Reduce the source impedance to 10K and you will have MUCH better luck. If you happen to be trying to measure a battery voltage and did not want to load down the battery then put an OPAMP buffer in between the A/D and the battery. Measuring battery voltage does not need to be so precise and any error introduced is not so big of deal.

I invite you to think about what happens in a circuit on a board that is designed with impedances in the 1 to 10 Meg range and then you go probing around in there with a 10M scope probe. Do you see on the scope the actual voltages in the circuit ?? (I hope you said "Of course not"). So if you use a seat of the pants rule of 100::1 to eliminate measurement error to acceptable levels then you would never use a 10M scope probe in circuits that had impedances over 100K or so. Right? You should agree with me. So if you have an A/D converter with an imput impedance of 100 MegOhms it is just on the margin of acceptability to use it in a 1 MOhm circuit. And if you are trying to use a high precision measurement circuit then the ratios should me more like 1000::1 for a 10->11 bit converter and 10000::1 for a 12->14 bit converter.

Good Luck
Michael Karas

BTW...It would help a lot if you could tell us what you are trying to measure and why you have it comfigured with a 1 MOhm source impedance,

MJK